Your daily source for AI news, tools reviews, and practical guides to artificial intelligence
Despite significant advancements, today’s leading AI models fail to perform simple visual recognition tasks that toddlers can accomplish effortlessly, exposing key limitations in current artificial intelligence systems.
OpenAI is reportedly developing an innovative AI-first audio wearable device codenamed ‘Sweetpea’ in partnership with renowned designer Jony Ive. This collaboration aims to redefine AI wearables with a focus on advanced audio capabilities.
Elon Musk has initiated a high-stakes legal battle against OpenAI and Microsoft, demanding up to $134 billion and questioning OpenAI’s nonprofit foundation amid growing scrutiny of the AI giant’s rapid commercial evolution.
OpenAI introduces the ‘Open Responses’ initiative aiming to unify AI application programming interfaces (APIs), potentially streamlining development and establishing its format as the industry benchmark.
Google has transitioned from a search-focused company to a leading AI powerhouse, advancing its AI models from Bard to the sophisticated Gemini series. This timeline highlights the milestones and capabilities that define Google’s AI journey.
Google is experimenting with a new AI-driven browsing feature named Auto Browse, powered by its Gemini assistant, aiming to make web navigation more intuitive and efficient within Chrome.
Mathematics genius Terence Tao acknowledges OpenAI’s GPT-5.2 Pro for solving an open Erdős problem autonomously, marking a notable achievement in AI capabilities while emphasizing that the result reflects computational speed over problem complexity.
Mathematics prodigy Terence Tao acknowledges OpenAI’s GPT-5.2 Pro for independently solving a longstanding open Erdős problem, marking a notable AI milestone. However, Tao advises caution in interpreting this achievement, suggesting that speed rather than problem difficulty is the key factor, with more significant AI advancements on the horizon.
Nicholas Moore has pleaded guilty to unlawfully accessing sensitive information from the Supreme Court and other federal agencies, sharing the data publicly on social media.
Nous Research has launched NousCoder-14B, an open-source AI model for competitive programming that rivals proprietary systems like Anthropic’s Claude Code. Trained in just four days on 48 Nvidia B200 GPUs, it achieves a 67.87% accuracy rate on LiveCodeBench v6, highlighting rapid advances and fierce competition in AI-assisted software development.